Industry View
Canvassing for Investors
Kevin Radell
01/01/2007

Kevin Radell is president of Predmore Holdings, a New York–based fine art and investment management consultancy. Predmore advises Artnet Worldwide on the development of market research products derived from Artnet’s signature Price Database.

The collapse of Fernwood Art Investments last summer because of "irreconcilable differences" rocked not only the art world, but also the mother of all potential patrons: Wall Street. Could personalities alone ruin such a tantalizing market opportunity, or are we missing something? For aspiring art fund managers, wearing art indices and market measurement tools on their sleeves can do more harm than good.

Imaginations had strained for nearly two years to understand why the self-proclaimed Merrill Lynch of art investments had not closed on at least one smallish art fund, despite Fernwood’s avowal that it had not shifted into fundraising mode. However, few would have guessed that internal discord would, or could, end this high-profile and unabashedly expensive American effort to tout art as the latest and greatest alternative asset among wealth managers and institutional investors.

Art had finally made it on the Wall Street radar screen as the darling new alternative asset class; last June, a single Gustav Klimt painting fetched a record $135 million.

In explaining the art industry from both a curatorial and financial perspective, no group exceeded Fernwood, which packaged its products accordingly. Masterful presentations to institutions and wealthy investors, integrating Nobel laureate Harry Markowitz’s efficient frontier model, showed how fine art reduces risk and increases expected return in a statistically measured investment portfolio. Like ABN Amro before it, Fernwood wore the mantel of the popular Mei Moses All Art Index on an as-needed basis to add impact. Proprietary stress testing in Fernwood presentations showed that art performs well during times of inflation, rising interest rates and even during down markets and periods of weaker economic growth.

Once convinced that art risk was quantifiable, investors could choose the Sector Fund or the Opportunity Fund, each capped at a proposed $100 million. Fernwood would manage the former vehicle from the viewpoint of an art collector, diversifying among the major collector categories while also providing a modicum of liquidity through structural gymnastics. The Opportunity Fund would operate as an art dealer, with an additional risk/return element and clearance to engage in other investment activities—such as short- and long-term lending and assisting with consignor guarantees. This product line is quite ingenious, as it suggests an insider’s understanding of the art industry articulated in the seemingly unassailable language of modern portfolio theory.

Personalities aside, why would any sane business organization blow such a story at a time when the white-hot art industry was enjoying its third year of a bull market? Art had finally made it on the Wall Street radar screen as the darling new alternative asset class; last June, a single Gustav Klimt painting fetched a record $135 million. One possibility concerns the dual impact of the questionable relevance of art data as an input for efficient frontier models, combined with the movement away from relative return portfolio management strategies among professional asset managers.

Relative Problems
Many believe that the art market is highly inefficient, thereby offering a window of opportunity to find undervalued situations. Reflecting this condition, art market data used as input for efficient frontier models intuitively suggests an oxymoron from the start. But let us dig deeper. Markowitz developed modern portfolio theory and the efficient frontier model in the 1950s. The latter uses advanced statistical techniques to measure the effect on portfolio risk and return of different securities, or asset categories, having different price movement correlations. For example, asset categories that have a relatively low level of correlation, such as real estate and stocks, can improve the level of expected return for a given level of risk when included in a diversified portfolio. The efficient frontier is the graphical curve representing the optimal mix of assets that will produce the highest return for each level of risk.

Efficient frontier models, explains money manager Edgar Weissenberger, assume that price data used for input is continuously observable, prices are fair and relate to homogeneous widgets, and transaction costs are low. Art data is poorly suited for such models because the inefficient art market has very low liquidity, little price transparency and steep transaction costs. Worse yet, art data is heterogeneous (i.e., no two widgets are alike) and occurs in seasonal data bursts, corresponding to the important spring and fall auction seasons each year. These attributes would seemingly provide little comfort to analysts scrutinizing the methodology used by Fernwood (or any other espouser of efficient frontier analysis for art investment) to make portfolio diversification recommendations.

When the tech bubble burst in 2000, causing Nasdaq to plummet and drag other indices into the vortex, few investors enjoyed immunity.

The other development in the investment industry that may have worked against broad acceptance of Fernwood’s offerings is the continuing shift among many portfolio managers away from relative return strategies and toward absolute return methods. Absolute return strategies were the hallmark of portfolio managers during the 1960s. The objective of an absolute return strategy is to generate a positive financial return for assets under management, regardless of broad market movements. Absolute return managers have wide discretion of when to buy or sell in any market of choice, a technique that most hedge fund managers today find effective.

As indices and portfolio management science gained popularity in the 1970s and early 1980s, institutional portfolio managers increasingly favored relative return strategies, attempting to show portfolio management prowess relative to popular indices in each asset class. These strategies stress quantitative theory and computer-driven statistical tools, such as the capital asset pricing model and related efficient frontier analysis, for the investment process.

Several factors in the capital markets around the turn of the millennium caused asset managers to reevaluate the effectiveness of relative portfolio management strategies. When the tech bubble burst in 2000, causing Nasdaq to plummet 50 percent and drag other indices into the vortex, few investors enjoyed immunity. Nor were they amused by merely losing 40 percent, albeit performing positively relative to the market indices. This incident, along with the seemingly inexorable trend toward correlation among traditional securities, shook the very foundation of the correlation matrix upon which relative strategies and efficient frontier models are based. Today, absolute returns are back in vogue, along with statistical coefficients such as alpha, a measurement of return attributed to specific securities independent of general market fluctuations. Investors increasingly look for fund managers who combine strong skills in fundamental analysis with the flexibility to move from market to market, unconstrained by benchmarks and indexing strategies.

Within this context, Fernwood’s pitch may have struck one or more wrong chords, despite the polish and seemingly sophisticated presentation. Institutions and experienced investors were in no mood to readily embrace an efficient frontier model fueled by erratic data from a new and often perplexing asset class in an inefficient industry. The irony of these circumstances is that an absolute strategy can work well in the inefficient art market. Indeed, the variety and dynamism of the many underlying category sectors provide fertile buying, selling and even arbitraging opportunities for nimble art fund managers. New products measuring market performance by artist and by collector categories serve to assist with investment analysis. The quality of art data, including the possibility of disclosed dealer prices and transaction costs, improves steadily. Meaningful price-performance indices for art sectors will surely follow.

The validity of art as investment remains, despite the collapse of Fernwood and ABN Amro. Aspiring fund managers may well benefit by speaking less of efficient frontiers and more about the benefits of taking an absolute return approach on an attractive—and temptingly inefficient—playing field.